Cross-plane color image encryption based on two-dimensional sine-henon map and genetic algorithm

In recent years, image encryption based on the permutation–diffusion architecture has been proposed, with the majority of them focusing on the pixel-level encryption. In this paper, a new image encryption scheme that combines bit-level encryption and pixel-level diffusion is proposed using a newly designed two-dimensional sine-henon chaotic map to generate pseudorandom sequences. The proposed bit-level encryption method considers the intrinsic characteristics of the image’s bit distributions. The high correlation among the bit planes is disrupted and shuffled using crossover and mutation techniques of genetic algorithms. Additionally, the nonlinear pixel-level diffusion is employed to enhance the avalanche effect of the image encryption process. Experimental results conducted on color images with sizes of 256 × 256, 512 × 512, and 768 × 512 demonstrate that the proposed scheme possesses a large key space of approximately 2199.32 and high sensitivity to slight changes in the secret keys. The encrypted images achieve information entropy values close to the ideal value of 8, NPCR values greater than 99.6094% and UACI values greater than 33.4635%, and low correlation coefficients close to zero in all directions, indicating strong randomness and excellent resistance against statistical and differential attacks. Therefore, the proposed encryption scheme provides a secure and effective approach for image encryption. © 2026 Elsevier B.V. All rights are reserved, including those for text and data mining, AI training, and similar technologies.